Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for ascertaining a piece of topological information of an intersection, the method comprising: locating a vehicle with lane accuracy when negotiating the intersection; ascertaining data by the vehicle when negotiating the intersection; transmitting the data to a processing unit; and ascertaining a connectivity of lane-roadway combinations of the intersection from the data with the processing unit; wherein lane change matrices are created from the data, which indicate from which lane-road combination the vehicle enters another lane-road combination, the lane change matrices being provided to a neural network as input data.
This invention relates to a system for determining topological information about road intersections, specifically the connectivity between lanes and roadways. The problem addressed is the lack of accurate, real-time data on how lanes at intersections connect to each other, which is critical for autonomous driving and navigation systems. The method involves a vehicle equipped with sensors and localization technology to precisely determine its position within an intersection with lane-level accuracy. As the vehicle navigates the intersection, it collects data about its movements, including lane changes and transitions between roadways. This data is transmitted to a processing unit, which analyzes it to map the connectivity of lane-roadway combinations. The system generates lane change matrices that represent how vehicles move from one lane-road combination to another. These matrices are then used as input for a neural network, which further processes the data to refine the intersection's topological model. The neural network can improve over time as more data is collected, enhancing the accuracy of intersection mapping. This approach enables dynamic, data-driven mapping of intersection layouts, supporting safer and more efficient autonomous vehicle navigation.
2. The method as recited in claim 1 , wherein locating the vehicle with lane accuracy is carried out using WGS85 coordinates.
3. The method as recited in claim 1 , wherein locating the vehicle with lane accuracy is carried out using street names.
4. The method as recited in claim 1 , wherein data of a defined high number of trips of the vehicles are used for ascertaining the connectivity.
This invention relates to a method for assessing connectivity in a vehicle network by analyzing trip data. The method involves collecting and processing data from a high number of vehicle trips to determine connectivity patterns. The system first identifies a set of vehicles equipped with communication devices and tracks their movement over multiple trips. Data from these trips, including communication events, signal strength, and network interactions, are aggregated to establish connectivity metrics. The method evaluates the frequency, duration, and reliability of connections between vehicles and network infrastructure. By analyzing a large dataset of trips, the system identifies recurring connectivity issues, optimal communication paths, and areas with poor signal coverage. The method may also incorporate environmental factors, such as terrain or weather, to refine connectivity assessments. The goal is to improve network reliability and performance by leveraging historical trip data to predict and mitigate connectivity problems. The system can be applied to autonomous vehicles, fleet management, or smart transportation networks to enhance communication efficiency and safety.
5. The method as recited in claim 1 , wherein a defined filtering of unlikely connectivity data is carried out during the ascertainment of the connectivity.
6. The method as recited in claim 1 , wherein the transmission of the data to the processing unit is carried out in real time.
7. The method as recited in claim 1 , wherein a neural network provides the data regarding connectivity in the form of a number or in the form of an adjacency matrix.
A system and method for analyzing connectivity data using neural networks. The technology addresses the challenge of efficiently representing and processing connectivity information in complex networks, such as social networks, biological systems, or infrastructure networks. Traditional methods often struggle with scalability and accuracy when dealing with large-scale or dynamic connectivity patterns. The invention involves a neural network that processes connectivity data and outputs it in a structured format, either as a numerical value or as an adjacency matrix. The neural network is trained to recognize patterns in connectivity, allowing it to generate a concise representation of relationships between nodes in a network. This representation can be used for further analysis, such as identifying key nodes, predicting network behavior, or optimizing network performance. The neural network may be configured to handle different types of connectivity data, including directed or undirected graphs, weighted or unweighted connections, and dynamic changes in connectivity over time. The output format—whether a single number or a matrix—depends on the specific application and the level of detail required. For example, a numerical output might suffice for a high-level summary, while an adjacency matrix provides a detailed breakdown of connections. This approach improves upon prior methods by leveraging machine learning to automate the extraction and representation of connectivity information, reducing manual effort and increasing accuracy. The system can be applied in various fields, including network security, transportation planning, and biological research, where understanding connectivity is critical.
8. A system for ascertaining a piece of topological information of an intersection, comprising: a locating unit configured to locate a vehicle with lane accuracy when negotiating the intersection; an ascertainment unit configured to ascertain data by the vehicle when negotiating the intersection; a transmission unit configured to transmit the data to a processing unit, the processing unit being configured to ascertain a connectivity of the lanes of the intersection from the data; wherein lane change matrices are created from the data, which indicate from which lane-road combination the vehicle enters another lane-road combination, the lane change matrices being provided to a neural network as input data.
9. A non-transitory computer-readable data carrier on which is stored a computer program, which is executable by a processor, comprising: a program code arrangement having including program code for ascertaining a piece of topological information of an intersection, by performing the following: locating a vehicle with lane accuracy when negotiating the intersection; ascertaining data by the vehicle when negotiating the intersection; transmitting the data to a processing unit; and ascertaining a connectivity of lane-roadway combinations of the intersection from the data with the processing unit; wherein lane change matrices are created from the data, which indicate from which lane-road combination the vehicle enters another lane-road combination, the lane change matrices being provided to a neural network as input data.
This invention relates to a system for analyzing and modeling traffic flow at intersections using vehicle data. The problem addressed is the lack of detailed topological information about intersections, which is critical for autonomous driving, traffic management, and navigation systems. The solution involves a computer program stored on a non-transitory data carrier that processes vehicle-collected data to generate lane-level connectivity maps of intersections. The system locates vehicles with high precision (lane accuracy) as they pass through intersections. Vehicles collect data, such as sensor readings or trajectory information, while navigating the intersection. This data is transmitted to a central processing unit, which analyzes it to determine the connectivity between lanes and roadways. The processing unit generates lane change matrices that represent how vehicles transition between different lane-road combinations. These matrices are then used as input for a neural network, enabling the system to learn and predict traffic patterns at intersections. The invention improves intersection modeling by leveraging real-world vehicle data, enhancing accuracy in mapping lane connections and supporting applications like autonomous vehicle navigation and traffic optimization. The neural network component allows for adaptive learning, improving predictions over time.
Unknown
March 16, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.